1. Field of the Invention
The present invention relates to the preparation and use of query trees, such as may be used in call centers, help desks, or informational applications and web pages.
2. Background of the Related Art
Help desks and call centers are set up to have personnel, referred to as call handlers, lead callers through a series of questions intended to first identify an individual or organization with the expertise to handle the call and then to determine a proper response to give the caller. Since call handlers may not be experts on the topics being discussed with the caller, a common approach is for the call handlers to use software that prompts them to ask specific questions of the caller while the software navigates a query tree. A “query tree” gets its name because a graphical representation of the series of questions connected by branches resembles a stereotypical triangular tree, with an initial node at the top branching downward to multiple nodes, which themselves branch to multiple nodes, until a much wider base is formed. Specifically, the call handler is presented with one question at a time to ask the caller, wherein the question corresponds to a node in the tree. Based upon the caller response to that question, a branching decision is made within the tree, leading to the next question or a reply to give the caller. This process of question and response can continue any number of times, as determined by a query tree designer or programmer, to navigate through the query tree. Theoretically, the series of questions and responses should eventually lead the caller to the information sought, which may be a voice connection to an individual with the required expertise, a recorded statement containing information about a given subject matter, or data leading to further information.
Rather than always involving a live call handler, some help desks or call centers may implement a fully automated query tree, such as an interactive voice response system. Typically, an interactive voice response system allows a caller to navigate a query tree by playing a prerecorded or voice synthesized question and then using voice recognition techniques to identify the caller's response. Ideally, the caller is able to navigate promptly and directly to the desired information. Unfortunately, a misunderstood question, a mistaken caller response, or an unrecognizable voice response can prevent the caller from obtaining the desired information.
A closely similar automated system requires the caller to enter numerical responses to each question, such as through a telephone keypad, in order to navigate the query tree. Although a numerical response is easier to recognize than a voice response, the question must generally include a statement explaining the meaning of several numerical response. These statements can become quite lengthy and difficult for the caller to remember and evaluate. For example, an automated system for a bank might say “Press 1 for new accounts; Press 2 for existing account balances; Press 3 for auto loans; Press 4 for hours of operation; Press 5 for retirement accounts; . . . ” A caller wanting information about setting up a new individual retirement account (IRA) may be confused whether to press 1 for new accounts or press 5 for retirement accounts.
Although the use of low-expertise call handlers and automated systems may result in significant financial savings to the businesses or governmental entities that employ them, there is potential for a caller to become frustrated at the amount of time and effort required to obtain the information that they need. A carefully designed and scripted query tree can significantly improve a caller's experience with such systems, but there remains a need for improved call systems.